ChangeLog suggestion

ChangeLog suggestion

[Zack Brown]

Hi Linus, Marcelo,

In each changelog entry, it would be really useful to include the
Message-ID of that email in a regex-parsable location. This way, if the
email was cced to lkml it would be possible for folks to track down the
actual patch.

I'm not familiar with your scripts, but I'd be surprised if this were very
difficult to implement. At the same time, there are many cases of changelog
entries that read only 'USB' or something equally unhelpful, where there is
little chance that anyone could track down the corresponding patch.  Having the
Message-ID in those cases would make all the difference in the world.

Be well,
Zack

-- 
Zack Brown

Re: ChangeLog suggestion

[John Bradford]

> In each changelog entry, it would be really useful to include the
> Message-ID of that email in a regex-parsable location. This way, if the
> email was cced to lkml it would be possible for folks to track down the
> actual patch.
> 
> I'm not familiar with your scripts, but I'd be surprised if this were very
> difficult to implement. At the same time, there are many cases of changelog
> entries that read only 'USB' or something equally unhelpful, where there is
> little chance that anyone could track down the corresponding patch.  Having the
> Message-ID in those cases would make all the difference in the world.

The changelogs are generated by BitKeeper - couldn't we simply include
a link that will let anybody[1] access the relevant changesets?

[1] This can be via HTTP, and _doesn't_ require anybody to use
BitKeeper in any way.

John.

Re: ChangeLog suggestion

[Zack Brown]

Hi John,

On Sat, Apr 26, 2003 at 07:52:19AM +0100, John Bradford wrote:
> > In each changelog entry, it would be really useful to include the
> > Message-ID of that email in a regex-parsable location. This way, if the
> > email was cced to lkml it would be possible for folks to track down the
> > actual patch.
> > 
> > I'm not familiar with your scripts, but I'd be surprised if this were very
> > difficult to implement. At the same time, there are many cases of changelog
> > entries that read only 'USB' or something equally unhelpful, where there is
> > little chance that anyone could track down the corresponding patch.  Having the
> > Message-ID in those cases would make all the difference in the world.
> 
> The changelogs are generated by BitKeeper - couldn't we simply include
> a link that will let anybody[1] access the relevant changesets?
> 
> [1] This can be via HTTP, and _doesn't_ require anybody to use
> BitKeeper in any way.

That would be ideal.

Be well,
Zack

> 
> John.
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

-- 
Zack Brown

Re: ChangeLog suggestion

[Linus Torvalds]

On Fri, 25 Apr 2003, Zack Brown wrote:
> 
> I'm not familiar with your scripts, but I'd be surprised if this were very
> difficult to implement.

Well, the scripts can take it, but quite frankly I'd rather not clutter 
the changelogs up with crud that really doesn't matter.

The thing is, the stuff that already _is_ in the changelog is certainly 
enough to identify the email if you just have a reasonable search engine. 
You have author, comments and diff, and if that isn't enough to identify 
the thing then something is wrong.

Also, _most_ of the patches by far end up coming as personal emails, and
while they have often shown up on linux-kernel in _some_ way, it won't be
the same email that got sent to me. The email that showed up on the
mailing list will often have been of the type "please test this out and if
it works for you I'll send it to Linus", or it will have been posted by
the original author and then the actual patch made it to me either through
somebody elses BK tree _or_ through a person like Andrew Morton or Alan
Cox.

In other words, what you ask for is ugly (yes, I actually look at the 
output of "bk changes") _and_ not very useful.

			Linus

Re: ChangeLog suggestion

[Linus Torvalds]

On Sat, 26 Apr 2003, John Bradford wrote:
> 
> The changelogs are generated by BitKeeper - couldn't we simply include
> a link that will let anybody[1] access the relevant changesets?

Well, yes, the changelogs are generated by BitKeeper, but what gets fed 
into bitkeeper is controlled by some scripts I wrote, which are the ones 
that take the email and munge it into a readable format etc. So by the 
time the thing hits my BK repository, the email headers will all have been 
thrown away, except for "From: " and "Subject: ". So BK never sees the 
full email.

(Even my scripts don't see the full email a large percentage of the time:  
I end up prettifying the emails for actual application by first removing
things like "Hi Linus, please apply this" etc which are pointless in the 
changelog).

		Linus

Re: ChangeLog suggestion

[Larry McVoy]

On Sat, Apr 26, 2003 at 09:34:25AM -0700, Linus Torvalds wrote:
> (Even my scripts don't see the full email a large percentage of the time:  
> I end up prettifying the emails for actual application by first removing
> things like "Hi Linus, please apply this" etc which are pointless in the 
> changelog).

I really wish Marcelo would do this or someone would volunteer to clean
up the change log comments.  BitKeeper doesn't have an easy way to 
propogate changes to the checkin comments (they are not currently 
revision controlled).  However, you can change them and someone could
clean up all the crud in the 2.4 tree and we could replace the one
on bkbits with a cleaned up one.
-- 
---
Larry McVoy              lm at bitmover.com          http://www.bitmover.com/lm

Re: ChangeLog suggestion

[John Bradford]

> > The changelogs are generated by BitKeeper - couldn't we simply include
> > a link that will let anybody[1] access the relevant changesets?
> 
> Well, yes, the changelogs are generated by BitKeeper, but what gets fed 
> into bitkeeper is controlled by some scripts I wrote, which are the ones 
> that take the email and munge it into a readable format etc. So by the 
> time the thing hits my BK repository, the email headers will all have been 
> thrown away, except for "From: " and "Subject: ". So BK never sees the 
> full email.
> 
> (Even my scripts don't see the full email a large percentage of the time:  
> I end up prettifying the emails for actual application by first removing
> things like "Hi Linus, please apply this" etc which are pointless in the 
> changelog).

Oh, I wasn't suggesting trying to preserve a link to the E-Mail via
the scripts->BK chain.

What I was thinking of was each time you import a patchset from a
[mail | set of mails], BK makes that in to a changeset, which can't be
referenced directy in the changelog, because the changeset's id can
change in the future.  Is there no way that BK can generate a
reference to the HTTP interface to bkbits that will remain constant,
and can therefore be included in the changelog, (as an URL to retrieve
a patch from).  Even if the patch can't be applied to any particular
tree, it's still something that can be refered to in a mail to LKML,
or similar.

John.

Re: ChangeLog suggestion

[John Bradford]

> > > The changelogs are generated by BitKeeper - couldn't we simply include
> > > a link that will let anybody[1] access the relevant changesets?
> > 
> > Well, yes, the changelogs are generated by BitKeeper, but what gets fed 
> > into bitkeeper is controlled by some scripts I wrote, which are the ones 
> > that take the email and munge it into a readable format etc. So by the 
> > time the thing hits my BK repository, the email headers will all have been 
> > thrown away, except for "From: " and "Subject: ". So BK never sees the 
> > full email.
> > 
> > (Even my scripts don't see the full email a large percentage of the time:  
> > I end up prettifying the emails for actual application by first removing
> > things like "Hi Linus, please apply this" etc which are pointless in the 
> > changelog).
> 
> Oh, I wasn't suggesting trying to preserve a link to the E-Mail via
> the scripts->BK chain.
> 
> What I was thinking of was each time you import a patchset from a
> [mail | set of mails], BK makes that in to a changeset, which can't be
> referenced directy in the changelog, because the changeset's id can
> change in the future.  Is there no way that BK can generate a
> reference to the HTTP interface to bkbits that will remain constant,
> and can therefore be included in the changelog, (as an URL to retrieve
> a patch from).  Even if the patch can't be applied to any particular
> tree, it's still something that can be refered to in a mail to LKML,
> or similar.

Ignore this suggestion, I missed your other mail in this thread - I
see now that there isn't a sensible way to do it :-).

John.

Re: ChangeLog suggestion

[Zack Brown]

Hi Linus,

On Sat, Apr 26, 2003 at 09:34:25AM -0700, Linus Torvalds wrote:
> 
> On Sat, 26 Apr 2003, John Bradford wrote:
> > 
> > The changelogs are generated by BitKeeper - couldn't we simply include
> > a link that will let anybody[1] access the relevant changesets?
> 
> Well, yes, the changelogs are generated by BitKeeper, but what gets fed 
> into bitkeeper is controlled by some scripts I wrote, which are the ones 
> that take the email and munge it into a readable format etc. So by the 
> time the thing hits my BK repository, the email headers will all have been 
> thrown away, except for "From: " and "Subject: ". So BK never sees the 
> full email.
> 
> (Even my scripts don't see the full email a large percentage of the time:  
> I end up prettifying the emails for actual application by first removing
> things like "Hi Linus, please apply this" etc which are pointless in the 
> changelog).

I think John's idea was that the email is not really important, as long as
the diffs themselves could be made available via URLs. That sounds good. I
only asked for the Message-ID because it didn't occur to me there might be
a straighter path to the diff.

I'm trying to deal with cases where the only changelog comment is
something like:

"kbuild: Hand merge link order change form driverfs update."

I know who did it. I know when it was accepted into the tree. I can spend
awhile grepping megabytes of lmkl archives (and come up empty, in this case),
but after that I have to give up. The patch is pretty much simply unavailable
at that point.

A URL to the diff itself would make it trivial to delve deeper into the
meaning behind changelog entries. Currently, investigating each entry just
takes way too much time, and often leads nowhere. If you can automate it
easily at your end, it would make a big difference to anyone studying
kernel development.

Be well,
Zack

> 
> 		Linus
> 

-- 
Zack Brown

Re: ChangeLog suggestion

[Arjan van de Ven]

[-- Attachment #1: Type: text/plain, Size: 1222 bytes --]

On Sat, 2003-04-26 at 17:12, Zack Brown wrote:
> Hi John,
> 
> On Sat, Apr 26, 2003 at 07:52:19AM +0100, John Bradford wrote:
> > > In each changelog entry, it would be really useful to include the
> > > Message-ID of that email in a regex-parsable location. This way, if the
> > > email was cced to lkml it would be possible for folks to track down the
> > > actual patch.
> > > 
> > > I'm not familiar with your scripts, but I'd be surprised if this were very
> > > difficult to implement. At the same time, there are many cases of changelog
> > > entries that read only 'USB' or something equally unhelpful, where there is
> > > little chance that anyone could track down the corresponding patch.  Having the
> > > Message-ID in those cases would make all the difference in the world.
> > 
> > The changelogs are generated by BitKeeper - couldn't we simply include
> > a link that will let anybody[1] access the relevant changesets?
> > 
> > [1] This can be via HTTP, and _doesn't_ require anybody to use
> > BitKeeper in any way.
> 
> That would be ideal.

short of that there's the bk commit mailinglist that has all of them
anyway; it's 3 clicks in evolution to find one given the author ;)

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: ChangeLog suggestion

[Shachar Shemesh]

Linus Torvalds wrote:

>The thing is, the stuff that already _is_ in the changelog is certainly 
>enough to identify the email if you just have a reasonable search engine. 
>You have author, comments and diff, and if that isn't enough to identify 
>the thing then something is wrong.
>  
>
Slightly OT:
Note that the "author" info is only semi readable from CVS. It contains 
just the user part of the email address (in my case - "lkml" for spam 
reasons - hardly a unique identifier). I'm told that under BK it has the 
full email, so I'm not sure where that stands there.

This does mean that under CVS, this info is not searchable, nor even 
guarenteed unique.

             Shachar

-- 
Shachar Shemesh
Open Source integration consultant
Home page & resume - http://www.shemesh.biz/

Re: ChangeLog suggestion

[Linus Torvalds]

On Sat, 26 Apr 2003, Shachar Shemesh wrote:
>
> Note that the "author" info is only semi readable from CVS. It contains 
> just the user part of the email address (in my case - "lkml" for spam 
> reasons - hardly a unique identifier). I'm told that under BK it has the 
> full email, so I'm not sure where that stands there.

People have indeed asked me to do that even for the BK tree - to avoid 
spammers picking up the address. I don't want to do it.

Personally, I want to have some address that I can reach the author at,
and if that means that spammers can pick it up too (open source means that
everybody has the same stuff _I_ have) I guess people who want to talk to
me need to live with spam filters or send patches to me from special 
accounts.

But both the short-format changelog (the ones posted to linux-kernel) and 
the CVS tree do hide the addresses somewhat.

			Linus

Re: ChangeLog suggestion

[Shachar Shemesh]

Linus Torvalds wrote:

>But both the short-format changelog (the ones posted to linux-kernel) and 
>the CVS tree do hide the addresses somewhat.
>
>			Linus
>  
>
I'm afraid I did not explain myself clearly. I fear that the format is 
both too short, and not descriptive enough. I think taking only the real 
name from the email would server both the purpose of searching the list, 
and the purpose of understanding who submitted the patch simpler.

The anti spam part was the fact that my email (for the purpose of this 
list) is lkml@shemesh.biz, which leaves an awfully undescriptive "lkml" 
in CVS. This allows neither searching nor accounting. As I'm sure other 
people use the exact same technique when posting to a mailing list, 
searching for someone who's email is lkml@* will not even be unique.

I hope my intention was clearer this time.

          Shachar

-- 
Shachar Shemesh
Open Source integration consultant
Home page & resume - http://www.shemesh.biz/

Re: ChangeLog suggestion

[Larry McVoy]

> The anti spam part was the fact that my email (for the purpose of this 
> list) is lkml@shemesh.biz, which leaves an awfully undescriptive "lkml" 
> in CVS. 

So fix your email, this is your problem, not the SCM's problem.  How hard
is it to change your email to shemesh-lkml or something reasonable?
-- 
---
Larry McVoy              lm at bitmover.com          http://www.bitmover.com/lm

Re: ChangeLog suggestion

[Jörn Engel]

On Sat, 26 April 2003 11:06:33 -0700, Linus Torvalds wrote:
> 
> Personally, I want to have some address that I can reach the author at,
> and if that means that spammers can pick it up too (open source means that
> everybody has the same stuff _I_ have) I guess people who want to talk to
> me need to live with spam filters or send patches to me from special 
> accounts.

It's not that bad. Since I started posting to lkml, my daily spam
dosis has risen from one to three, roughly. Annoying, but not a big
problem, even without filters.

Jörn

-- 
Victory in war is not repetitious.
-- Sun Tzu

Re: ChangeLog suggestion

[b_adlakha]

On Saturday 26 April 2003 23:59, Jörn Engel wrote:
> On Sat, 26 April 2003 11:06:33 -0700, Linus Torvalds wrote:
> > Personally, I want to have some address that I can reach the author at,
> > and if that means that spammers can pick it up too (open source means
> > that everybody has the same stuff _I_ have) I guess people who want to
> > talk to me need to live with spam filters or send patches to me from
> > special accounts. 
>
> It's not that bad. Since I started posting to lkml, my daily spam
> dosis has risen from one to three, roughly. Annoying, but not a big
> problem, even without filters. 
>
> Jörn

All the spam comes because of the archives of this mailing list stored as 
HTML
in thousands of places. The web-spider programs used by the spammers are 
sure
to find atleast one.
My inbox is full of stuff like "Strictly Confidental" and "your solution to 
never ending spam"